Piksel expects AI to be deeply involved in many aspects of the Media Supply Chain in the future.
Artificial Intelligence (AI) is suddenly all the rage, with the potential to assist in many areas of the industry from network diagnostics to edit assembly. The Broadcast Bridge gets under the skin of AI to ask how useful is machine learning today in media today. Kristan Bullett, Head of Solutions, Piksel shares his thoughts.
How can AI and machine learning be used to help streamline operations or aid better decision making?
Kristan Bullett (Piksel): The internet was designed to route traffic efficiently, but the sheer volume of data it deals with now results in time-expensive distribution issues. Analysis conducted by AI and machine learning will be able to support a clearer understanding of how to more efficiently route traffic and get data closer to MVPD. This will enable service providers to both reduce time to market and costs across distribution.
How can AI/ML impact the creative or editorial process of a production today?
KB: We see AI/ML supporting the editorial aspects of production in many ways. AI and ML are able to conduct automated inspection of content to provide deeper metadata identification and linkage. Machine learning can be used to match shows and movies with a greater than 95% accuracy, so service providers can use this improved consistency to give their customers better search and UI provision. Machine learning can also assist with matching to third party metadata providers to further improve accuracy. Additional uses for AI and ML include:
- Provide quality control “tips” and identify potential QC issues that can be checked by human editors
- Automated identification of blackframes and scene change
- Scene level analysis as a contribution to production
What are the current limits of AI/ML in media?
KB: The field is advancing very rapidly. Entity identification is an example of a relatively well established “real” ML technique, and, whilst the results need to be improved this is not about data analysis. Companies should remember that comparing ML with well applied data analysis is not of much value. What the machine does vs how it does it is not as important as the results that are achieved by the process. ML/AI will replace a lot of the manual efforts previously required and will therefore enable breathing space, from a creative perspective, or empower operational teams to handle more content with less people.
What should (and are) media companies be doing to incorporate AI/ML into their operations?
KB: The important point for businesses to remember is not to do all at once. Media organisations will take some time to adjust to ML/AI and I believe that the data output from ML/AI will, initially, compliment organisations. I fully expect some organisations will be very cautious at the prospect of machines “freeing up” staff time as the finance team evaluate cost saving options in this space. However, ML/AI does not need to be about resource reduction and can be used to provide powerful tools for operators looking to make better use of data – for example segmentation information or data to improve quality control.
What will AI be able to do by 2022 that it can’t do now?
KB: We expect AI to be deeply involved in many aspects of the Media Supply Chain in the near future but there needs to be caution about over exaggerating features. I expect AI/ML to support extensive automation of the Media Supply Chain from commission all the way through to distribution. Having the capability to intelligently distribute content efficiently will support further cost and latency reductions across the chain.
You might also like...
In its essence, a Media Management system is not built to generate revenue for organisations. The implementation of one can, however, streamline workflows and free-up time for revenue-generating activities. Of course, every organisation is different and so that means what…
Considering the unarguably fast clock speed in Media & Entertainment (M&E) today, content enterprises need to be able to rapidly access, preview, share, process and publish content for on-time delivery to an ever-increasing number of platforms and devices.…
Virtualisation is being touted by some vendors and consultants as the ultimate solution to handling media. Yet, for many engineers, it remains an untested answer to their production and playout needs. Here are two articles that provide important insight into…
Tuesdays HPA Tech Retreat was all about 360 and VR, and Wednesday focused on the versioning explosion. On the final day, delegates were given a summary of the current state of the industry, and the influences of artificial intelligence on media…
Yesterday’s 2017 HPA Tech Retreat in Oxford, UK, was all about VR and 360, and on Wednesday, they moved to the thorny issue of the versioning explosion. As broadcasters seek wider audiences over different platforms, localisation has become a big issue.…